Metric Learning Improves the Ability of Combinatorial Coverage Metrics to Anticipate Classification Error.
Tyler CodyLaura J. FreemanPublished in: ICSTW (2023)
Keyphrases
- classification error
- metric learning
- mahalanobis metric
- distance metric
- training set
- training samples
- multi task
- decision rules
- pairwise
- semi supervised
- feature space
- learning tasks
- generalization error
- class labels
- dimensionality reduction
- covariance matrix
- distance function
- hyperparameters
- base classifiers
- computer vision
- semi supervised learning
- estimation error
- cross validation
- transfer learning
- high dimensional data
- text mining
- upper bound
- image processing